Embed Size (px)
Citation preview
OPTIMISING ALTERNATIVE FUEL COMBUSTION AND BENEFITS:
A CASE STUDY (PART 1)
How FCT's modelling and process expertise helped a European cement plant improve their alternative fuel
firing substitution rate
Robert Jansky, Cemmac S.A.; Constantine Manias, FCT-Combustion Pty Ltd;
Roger Hassold, FCT-Combustion Pty Ltd
Customer: Cemmac, Horne Srnie, SlovakiaIndustry: CementKiln type: 5-stage ILC kiln (3.4m x 46m long)Fuel: Residue Derived Fuel (RDF) and coal
ABOUT THE CLIENTCemmac operates a 5-stage ILC kiln (3.4m x 46m long), which produces 1,200t/d of Portland clinker.
At the time of the study, coal and Residue Derived Fuel (RDF) were fired in the kiln by a 30MW burner from another supplier based in Europe. Additionally, 2 calciner burners, from that same supplier, were used to fire coal in the calciner. Tyres were also fed to the kiln inlet and RDF to the calciner at the twin tertiary air inlets.
THE SITUATIONCemmac approached FCT Combustion as they wished to increase their alternative fuel substitution rate.
The rate of RDF firing in the calciner was limited to 20% due to high CO levels, measured at the exit stage 4 cyclone (with stage 5 being the bottom stage) in excess of 0.3% or 3,000ppm. Even when firing 100% coal, the firing rate was limited by CO which in turn limited the degree of calcination and overall plant capacity (LOI of the stage 5 hot meal high at 8.4%).
Severe problems with build-up forming in the area above the kiln inlet chamber were also identified. To remove the build-up, the kiln was stopped for several hours roughly once every two weeks.
A process and combustion study of the kiln, as well as a more detailed Computational Fluid Dynamics (CFD) study of the calciner, were undertaken to:• Improve coal combustion• Improve RDF combustion• Reduce CO levels in kiln inlet• Increase alternate fuel substitution• Consider introducing a new alternative fuel, Polyurethane (PUR) dust, to calciner• Reducing build-up in the kiln system• Improve operational stability
PROCESS EVALUATIONSProcess evaluations of sulphur and chloride build-up in the system can be seen in Figure 1 and Figure 2.
For a kiln with favourable operating conditions, the volatility of SO3 in the kiln would normally be in the range 0.3 to 0.5, but Cemmac was averaging around 0.85.
WHY WAS SO3 VOLATILITY (φ) SO HIGH?The volatility of SO3 is dependent on the form that the SO3 is in (K2SO4, Na2SO4 or CaSO4) and the operating conditions of the kiln and increases with:
• An increase in localised reducing conditions such as flame impingement and fuel in the clinker bed• A reduction in oxygen levels in the kiln (giving global reducing conditions)• An increase in the sulfur/alkali balance above 1 (in the kiln system inputs)• A reduction in the clinker granule size (easier diffusion to the granule surface)• An increase in the maximum burning zone temperature• An increase in the retention time in the burning zone
FCT was established more than 35 years ago to bring more science into the art of industrial combustion.
Since then we have gained a reputation for technical excellence and introducing innovative, value-adding technologies for high temperature processing industries.
Kiln and calciner burner systems have a major impact on the bottom-line results for a cement plant, and so customised optimisation of the flame for any given kiln is critical to get the best plant performance.
We are combustion experts with a strong history of using various types of modelling to optimise burner designs and deliver performance.
1 2
FIGURE 1PROCESS EVALUATIONS
Sulphur and Chloride build-up in system
Red lines indicate normal volatility
range
SO3,CL
SO3,HM
φSO3 = 1 -
FIGURE 2PROCESS EVALUATIONS
Sulphur and Chloride build-up in system
FIGURE 3PROCESS EVALUATIONS
Sulphur/Alkali ratios
HOT MEAL - SULFUR BALANCE
[SO3/80] / K2O/94+Na2O/62-cl/71]
DESIRABLE RANGE (0.7 - 1.2)
FIGURE 4PROCESS EVALUATIONS
Sulphur/Alkali ratios
SULFUR BALANCE (SO3/80) /
(K2O/94+Na2O/62-Cl/71)
3
FIGURE 5CHEMICAL FACTORS
FIGURE 6CHEMICAL FACTORS
FIGURE 7CHEMICAL FACTORS
Clinker burnability - LSF vs SR
FIGURE 8CLINKER PROPERTIES
Clinker %Liquid Phase or %Flux at
1450ºC
4
FIGURE 9CLINKER PROPERTIES
FIGURE 10CLINKER GRANULOMETRY
FIGURE 11UNBURNED FUEL LANDING ON
THE CLINKER BED
5
WHY IS %CI SO HIGH IN HOT MEAL?Cl is high in hot meal because the input is higher than the output, accumulating in the kiln at a rate of approximately 18kg/h. Cl input from the raw meal, coal, tyres and RDF as a percentage of the raw meal is 0.098%.The Cemmac kiln has a chloride bypass for removal and control of chlorides within the kiln. Cemmac bypass is only sized for ~4%, however the rough 'rule of thumb' for sizing of a Chloride bypass is: %Kiln Exit Gas Bypassed = %Cl of the raw meal on LOI free basis x 100.%Bypass Required for Cemmac = %0.098 x 100 = 9.8%.
CONCLUSIONS FROM PROCESS STUDY• There are excessive amounts of sulphur and chloride in the hot meal that will lead to build up and operational instability.• There is excessive volatilisation of sulphur from the clinker that leads to this.• There is nothing unusual with the clinker chemistry or granulometry that gives rise to this – sulphur/alkali ratio of the clinker is quite low.• The burner momentum is quite high – so increasing burner momentum will not improve the retention of sulphur in clinker.• AF fuel particles are falling on the clinker bed and most likely the cause of the sulphur volatility, but there is probably a flame shape issue as well. • Adding Gypsum into the raw meal is making the situation worse – why is gypsum added?• The chloride bypass in undersized for the amount of chlorine in the system and should be increase to 10% to cope with current levels of chlorine.
FIGURE 12:
APPROXIMATE CHLORIDE MASS BALANCE FOR CEMMAC KILN
6
CALCINER & CFD MODELLING
The rate of combustion of each solid fuel is controlled by:
• The chemical composition of the fuel itself,• Particle size and shape, and• The aerodynamic trajectories that the fuel particles follow.
REQUIREMENTS TO ACHIEVE BURN OUT OF SOLID ALTERNATIVE FUELS IN CALCINER
TEMPERATURE
OXYGEN/ MIXING
TIME
SIZE/ SHAPE
BURN OUT
TEMPERATURE
• ↑Temp = ↑Rate of AF combustion
• CaCO3 → CaO+CO2 =
↓Temp
SIZE/SHAPE
↑Rate of combustion &
↓Retention Time by:
• ↓AF size
• AF 3D → 2D
TIME
Calciner retention time ↑by: • ↓AF size • AF 3D →2D • ↓Gas velocities • ↑Calciner volume
OXYGEN/ MIXING
Rate of AF combustion↑ by: • ↑O2
• ↑Turbulence/ Mixing
7
CFD CALCINER AERODYNAMICS KILN FLUE GAS AND AIR FLOW PATTERNS(EXISTING CONFIGURATION)
Some of the flue gas loops around in a small diameter
Some of the flue gas loops around in a large diameter
Some of the flue gas bypasses the mixing chamber
8
CFD COAL COMBUSTION: POOR COAL DISPERSION(EXISTING CONFIGURATION)
L4 LHS Burner L3 RHS Burner
L4 LHS Burner L3 RHS Burner
Coal particle traces coloured by volatile mass fraction.
Mostly greenMostly yellow
Good dispersion
Poor dispersion
Char combustion
starting before the
mixing chamber (blue
= very little char mass fraction)
Char combustion starting in the mixing
chamber (blue = very little char mass fraction)
9
CFD COAL COMBUSTION: IMPROVED COAL DISPERSION & COMBUSTION(EXISTING CONFIGURATION WAS MODELLED VERSUS PROPOSED CONFIGURATION)
Coal particle traces coloured by volatile mass fraction.
Relocated burner from front to rear Both burners to Level 3
Relocated burnersExisting burners
Coal char only begins to burn out in the mixing chamber
Improved coal char combustion commencing lower in the calciner
10
CFD COAL COMBUSTION: IMPROVED COAL DISPERSION & COMBUSTION(EXISTING CONFIGURATION WAS MODELLED VERSUS PROPOSED CONFIGURATION)
RDF COMBUSTION: SORTING INTO HEATING VALUE, PARTICLE SIZE & SHAPE GROUPS
FUEL GROUP 1: LHV= 13,240kJ/kg
cf. 30, 650 kJ/kg Coal
a. 2D Paper; b. 3D Wood
c. Close up of 3D wood. Even for a small wood
particle (size indicated in red) the burn out time in
typical calciner conditions is in the order of 15-30
seconds.
d. Experimental burn out time for a single pine
wood cylinder 1.65mm diameter x 6.6mm long in an air velocity of 1.5m/s
REMAINING UNSORTED RDF
k. +2mml. -2mm
FUEL GROUP 2:LHV= 23,520kJ/kg
cf. 30,650 kJ/kg Coal
e. & f. 2D thin film and thick
plastics
g., h., i. & j.3D plastic, rubber &
foam
a.
l.k.
j.
i.
h.
g.
f.
e.
d.c.
b.
11
CFD COAL & RDF COMBUSTION: RDF 3D PARTICLE SIZE VS RESIDENCE TIME
• Adjust quickly to the gas flow• Follow the kiln flue gas flow path• 2-8 second residence time• Burnout time of many of these particles is longer than the residence time in the calciner
• Majority fall into kiln• Those that exit with the flue gas have ~24 second residence time• Burnout time greater than residence time
• Bounce around more• Some fall into kiln and have very long residence times• Those that exit with the flue gas have 8-24 second residence time• Burnout time greater than residence time
Diameter <0.5cm2-8 seconds
Diameter >2cm
Diameter 1-2cm
12
CFD COAL & RDF COMBUSTION: RDF 3D PARTICLE SIZE VS RESIDENCE TIME CFD COAL & RDF COMBUSTION: RDF COMBUSTION PROBLEM SUMMARY
CONCLUSIONS FROM THE CALCINER AND CFD MODELLINGDue to the relatively large size of the RDF (cf Coal), the burn-out time was much longer than the residence time of the calciner.
The calciner is designed for coal combustion and does not have a sufficiently high solids:gas residence time ratio for combustion of the RDF within the calciner. In addition, the calciner design does not provide a high temperature zone for combustion of difficult fuels.
As a result, unburnt volatiles and char would exit the calciner and continue to burn, emitting CO which was then detected by the gas analysis system, limiting the fuel firing rate. In addition, unburnt char was collected in the bottom stage cyclone and entering the kiln with the hot meal, causing reducing conditions.
Whilst relocating the coal burners improved coal combustion, it made little improvement on RDF combustion. A number of changes were made to the CFD model to try and address issues such as raising the temperature and improving the heating value, but none made sufficient improvement to justify the change. The changes included:
• Increasing the tertiary air temperature by 100ºC (from 950 to 1050ºC).• Raising the meal feed pipe entry point to produce a high temperature zone below the meal entry point and accelerate RDF and coal combustion.• Reducing the RDF moisture (from 17.5 to 10%) to reduce the drying time and improve the LHV.
As an example, raising the meal entry point had the desired effect of raising the temperature but was still insufficient to ensure burn-out of the RDF.
Alternate raised meal entry point
Existing meal entry point
Low temperature with all meal to existing point
High temperature with all meal to upper entry point
In both cases RDF char mass fraction in the calciner flue gas is very high
13
CFD DUSTY ALTERNATIVE FUEL COMBUSTIONAs discussed, the biggest problem with the RDF was its large size. A new dusty alternative fuel injected through the coal burners was modelled. The alternative fuel dust has both a very fine particle size and good heating value.
Three alternative fuel dust firing cases were modelled:
• 100% replacement (3.25t/h PUR)
• 36% replacement with the remainder 64% coal (1.26t/h PUR and 1.64t/h coal); and
• 36% replacement with the remainder coal 46% and RDF 18% (1.26t/h PUR, 1.17t/h coal and 0.8t/h RDF)
In all cases, alternative fuel dust and coal combustion was good but there was no change to RDF combustion. With the use of alternative fuel dust, it is important that the total input of Chlorine to the kiln has been considered.
CFD COAL & RDF COMBUSTION: RDF COMBUSTION PROBLEM SUMMARY
AF Dust particle traces coloured by volatile mass fraction
AF Dust particle traces coloured by char mass fraction
The small amount of char is all but completely burnt out by the time it
leaves the mixing chamber
The AF is made up of almost entirely volatiles which are liberated very
quickly
14
CONCLUSIONS FROM CFD STUDYRe-location of the coal burners to the low velocity region on level 3 would provide a marked improvement in coal combustion in the calciner. There is also improved combustion from smaller 2D particles of the various alternative fuels used. However, the 3D particle combustion is not improved significantly by a relocation of the burners - the particle shape/size is the problem.
Multiple burners, increasing the tertiary air temperature, reducing alternative fuel moisture, and creating a high temperature zone were also modelled and likewise provide only a marginal improvement in the combustion of alternative fuel 3D particles. For improved combustion of these particles, either the calciner residence time must be increased, the alternative fuel particles must be limited to 1mm maximum in any dimesion, or some other novel design be considered that allows more time for particle burn-out.
RECOMMENDATIONS• Much of the sulphur volatilisation was arising from the kiln burner flame shape and impingement. It was recommended that this be addressed so that the sulphur content of clinker can be increased to normal levels, as well as to reduce the build-up being experienced.
• Calciner burners should be relocated to the level 3 low velocity region of the calciner.
• The practice of adding gypsum to the raw meal should be stopped if possible.• The chloride bypass should be increased to 10% (from 4%) of kiln gases.
• The alternative fuel dust is a good option and as much of the RDF as possible should be replaced with this.
• Reduce RDF particle sizing further if possible.
• Consider calciner enlargement options to increase residence time.
Stay tuned for the next edition of the FCT Combustion newsletter, 'Burn After Reading', where we will include the next part of this case study addressing burner design and selection.
CONCLUSIONS AND RECOMMENDATIONS
15
Asia-PacificFCT CombustionT +61 8 8352 9999E [email protected]
U.S.A & CanadaFCT Combustion IncT +1 954 866 0788E [email protected]
EuropeFCT Combustion GmbHT +49 3 222 109 6283E [email protected]
South AmericaFCT Combustao Brasil LtdaT +55 11 4118 6004E [email protected]
